Paste Japanese text above and the counts update as you type. Nothing is uploaded — the analysis runs in your browser.
"How long is this?" has several answers
For English, length is a word count and everyone means the same thing. For Japanese, the question splits:
- A form with a character limit counts every character, spaces included.
- A translator quoting for work counts source characters, usually excluding spaces.
- A learner checking difficulty wants to know how much of the passage is kanji, because that is what governs how hard it is to read.
- A designer wants the count that predicts how much space the text needs.
So this reports the breakdown rather than one number. The one you want depends on why you asked.
Why the word count is not the answer
Japanese does not put spaces between words. 私はネコが好きです is one unbroken run of characters containing roughly five words, and finding the boundaries requires a dictionary and a morphological analyser — the software has to know that 好き is a word and that は is a particle attached to what precedes it.
This tool counts whitespace-delimited runs, which for Japanese prose means it will report 1. That is deliberately literal rather than a guess: a word count that is confidently wrong is worse than one that is obviously not applicable. For mixed text with Latin in it, the number becomes meaningful again.
What counts as what
- Hiragana — the U+3041–U+309F block.
- Katakana — the full-width block and half-width katakana, since ｱ and ア are the same syllable at different widths.
- Kanji — every Unicode block holding ideographs used in Japanese, including the supplementary ranges. Without those, rare kanji such as 𠮟 and 𩸽 and compatibility characters such as 﨑 (common in surnames) fall through and get miscounted as Latin letters.
- Latin letters — including full-width Ａ–ｚ, which are still letters.
- Digits, whitespace, other — the remainder, with punctuation and symbols in "other".
The iteration mark 々 counts as kanji, because it stands in for one. ヶ and ヵ count as katakana, because that is the block they live in and how a reader encounters them.
Characters are counted by code point
A kanji outside the basic Unicode range is stored as two units internally, and a naive counter reports it as two characters. 𠮟 is one character; this counts it as one. If you have ever had a character limit reject text that looked short enough, this is often why.
Line counts follow what a text editor shows: an empty box is one line, a trailing newline opens another, and a Windows line break counts once rather than twice.
When you would use this
- Quoting or invoicing translation work priced per source character.
- Fitting a form or a field with a character limit.
- Judging reading difficulty from the proportion of kanji.
- Checking a text is what you think it is — a high "other" count usually means stray formatting characters.